There’s a really interesting piece on the Thesis blog about a split test on a newsletter subscription form. It’s worth checking out the full post and the comments, but in essence the test was around three variations of a simple email subscription form: Far left is control, the middle version tests without the ‘social proof’ [...]
